sales
Warn
Audited by Snyk on Feb 19, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.90). This skill's SKILL.md explicitly instructs sourcing and using user-generated public content (e.g., "Search subreddits", "Look at who's posting", "I saw your G2 review", LinkedIn posts, Product Hunt) as "pain signals" to personalize outreach, meaning the agent is expected to read and act on untrusted third-party content from open web sources.
Audit Metadata